410 



Fishery Bulletin 100(3) 



16 



12 



A S.E. LSI 



^^" Surface 

 I I Subsurface 



i 



i 



ii 



xu 



B So. Bock 



j/- jj2f", j^*'. i^*- ,6*^- j^*- X,*- .-z!" v*- .*-_ j^ .->- jt!" j^- .-^^ .^>^ 

 ^ s? 9P <-■?• '^^ 



# 



y>yt^^'$<«^^5^^^5/- 



■6" 



.^" 



Family 



Figure 3 



Abundance (mean no. larvae per 1000 m'l of 16 families of lar\'ae that 

 constituted at least f^r of the catch by surface and subsurface channel 

 nets at lA) S.E. LSI (southeast Lee Stocking Island.), and (B; So. Bock 

 (South Bock Cay). Error bars are 1 SE. 



cantly more fish larvae than hght traps anchored to the 

 sea floor, apparently independent of taxa. Thorrold noted 

 that this pattern was unexpected because more water 

 (and presumably more larvae) should pass by anchored 

 traps, and he suggested that the ability of larvae to swim 

 to and enter anchored traps may be difficult under high- 

 current conditions. 



Although we conclude that between-site differences in 

 relative catch by light traps and channel nets in our study 

 were related to current velocity, the relationship between 

 catch and average current velocity was not linear. The pro- 

 portional abundance of larvae caught in light traps was 



over 31 times higher at S.E. LSI than at So. Bock, whereas 

 the larval catch in channel nets was only 1.3 times higher 

 at So. Bock than at S.E. LSI. The nonlinear relationship be- 

 tween current velocity and larval abundance may represent 

 a threshold response in which the efficiency of light traps 

 and channel nets may change with different current veloc- 

 ities. The mechanisms causing such relationships might 

 be that fish larvae are able to orient to and swim into light 

 traps more easily under lower current velocities, whereas 

 fish larvae might more easily detect and avoid chan- 

 nel nets because of greater hydrodynamic disturbances 

 in front of nets during higher current velocities. 



